pieces analysed
860,000Sixteen months of sales and production records. We matched 92% between the two.
The profit brain for fresh food.
Eclipsai finds where fresh-food operations lose profit, makes the change, and proves the impact in cash. It starts with production, then extends to orders, pricing, buying, and labour.
The till records what sold. It does not record what was left, what sold out, or what customers asked for after the shelf was empty. That is where profit is lost.
What owners know and systems miss
What cannot stay on the shelf goes into surprise bags or the bin. The count is rarely recorded.
Tomorrow's order is set from averages, templates, special orders, and experience while the shop still needs cleaning.
Yesterday's decision becomes today's perishable stock.
A sell-out may mean the plan was right, or it may mean a missed sale. The till cannot tell which.
The till records what sold. It does not record what was left, what sold out, or what customers asked for after the shelf was empty. That is where profit is lost.
The profit brain at work
Eclipsai connects sales, production, ordering, and invoicing systems with email, team chats, and relevant external data. When the data cannot explain what happened, it asks staff a targeted question. They can answer with text, photos, or voice notes.
How Eclipsai works
A multi-site fresh-food operator gave us sixteen months of records. We connected its POS and production data and matched what was delivered with what sold across shops, products, and days.
Sixteen months of sales and production records. We matched 92% between the two.
The pattern concentrated in particular shops, products, and weekdays.
Recurring production patterns no longer matched demand. The ingredient cost of all unsold pieces was €190K.
Ingredient savings exceeded the full margin of any sale the cut might have missed. The cuts were net positive in all nine months tested.
Some waste protects sales. The €40–60K opportunity came from repeated overproduction after demand had changed.
From one multi-site operator's records, 2024 to 2025. Specific to that business, not a promise.
In our historical replay, letting one forecast rule set every production order lost money. It cut waste, but small forecast misses became sell-outs whose lost margin outweighed the ingredients saved.
Low-volume demand is noisy. The record does not contain every local event, or tell you what a sell-out meant.
Eclipsai adds the missing information, makes only the few changes the evidence supports, and measures the impact in cash.
Beyond production
Once connected to the company's systems and communication channels, Eclipsai can add the information needed for each new decision, implement the change, and measure the result.
Protect sales while cutting repeated, avoidable waste. Update production plans that no longer match demand.
Flag supplier price rises and prices that no longer cover costs.
See when smaller batches save waste but add work, or when understaffing costs sales.
Use what works in your current shops to start the next one.
It connects sales, production, and what your team sees to find the few decisions worth acting on and prove the result in cash.
For growing fresh-food operators with 2 to 20 locations.
Book a 20-minute callCommon questions
Eclipsai weighs both risks. A missed sale can cost more than the ingredients saved. It only suggests making less when the evidence supports it, then measures the result against the plan it replaced.
Eclipsai keeps the current plan as the baseline, adds what the records miss, and shows only where the evidence supports a different order. Once approved, it can implement the change in the production system and measure the result in cash.
Eclipsai works from sales and production records. It asks the team a short question only when the records cannot explain what happened, such as a sell-out, unusual leftovers, or a local event.
We match the records that can be trusted and make the gaps visible. If the data cannot support a recommendation, Eclipsai asks for context or leaves the plan alone.
No. We begin with the shop-product combinations where the evidence is strongest. We use the result there before applying the same rule elsewhere.
Starting is free. If you continue, we charge a monthly subscription based on the number of locations and decision areas monitored.